Amid the AI gold rush, a key realization is emerging: Extremes aren’t always sustainable. Businesses chasing moonshots, like building all-knowing chatbots or industry-transforming solutions, often face lengthy timelines and high failure rates. Meanwhile, over-automating repetitive tasks — such as deploying GenAI agents across isolated functions to handle small-scale processes like ticket resolution or basic queries — delivers limited, diminishing returns.
What’s missing? The forgotten middle — operational, mid-tier AI use cases that can solve specific pain points, scale seamlessly, and drive return on investment (ROI) faster than either extreme.
But there’s a bigger picture: ROI from AI isn’t a one-time breakthrough. The future belongs to enterprises that treat ROI as cumulative, built through a network of interconnected, operational AI initiatives. Achieving this requires not just a new lens on AI investments but also a cultural and organizational shift toward operationalization.
What Is the Forgotten Middle?
Businesses often approach AI through two ends of the spectrum:
- Moonshot AI Projects: Ambitious, high-risk initiatives like the all-knowing chatbot — an always-on, always-accurate assistant that delivers real-time business insights and future predictions — or industry-wide transformations. These require substantial R&D investment, long timelines, and uncertain ROI. For a more grounded approach to driving ROI, explore these three concrete ways to drive AI ROI.
- Low-Level Automations: Simple, repetitive tasks automated by rule-based systems or basic AI, like invoice processing. These automations are quick to implement and save significant time, especially for businesses managing large volumes. However, their utility often remains limited to isolated use cases. They typically lack scalability, as adapting them to new workflows or integrating them into broader systems can require additional effort. Without a cohesive strategy, low-level automations may create siloed efficiencies, which, while valuable in specific contexts, fail to contribute to enterprise-wide transformation.
The Forgotten Middle focuses on operational, mid-tier AI use cases: scalable projects embedded in workflows that solve specific, tangible business problems. These initiatives strike a balance between agility and impact, delivering faster ROI with lower risk. Examples include:
- Fraud Detection: Modern fraud detection uses AI to uncover complex patterns like synthetic identity fraud and cross-border transaction anomalies. These scalable solutions reduce false positives, speed up detection, cut costs, and minimize financial risk.
- Demand Forecasting: Advanced AI enhances traditional demand forecasting with greater precision and scalability. By integrating real-time data — such as promotions, weather, and social trends — businesses can optimize inventory, reduce waste, and prevent stockouts. These use cases deliver faster ROI and scale seamlessly across teams and markets.
- Hyper-Personalized Marketing: AI-powered insights enable businesses to craft hyper-targeted campaigns that boost engagement and ROI. By analyzing behaviors, preferences, and purchase patterns, companies can deliver tailored offers, personalized recommendations, and precisely timed campaigns. These strategies drive higher conversions, deepen customer loyalty, and create a competitive edge.
These mid-tier projects combine practicality with measurable outcomes, making them indispensable for long-term AI success.
The Case for Mid-Tier AI Use Cases
1. Less Glamorous, More Impactful
Why are mid-tier use cases undervalued? Because they lack the allure of moonshots. However, these projects are embedded in real-world workflows and closely align with measurable business outcomes.
For instance, predictive inventory management doesn’t grab headlines like autonomous warehouses but generates significant ROI by reducing waste and improving stock turnover. Platforms like Dataiku allow businesses to scale these use cases rapidly while mitigating the risks of jumping straight to moonshots. By building on foundational projects — starting with descriptive analytics, progressing to predictive use cases, and incorporating generative AI — organizations can confidently pave the way toward ambitious, transformational goals like agentic workflows.
2. Cumulative, Enterprise-Wide ROI
AI’s true value lies in the cumulative effect of interconnected initiatives that collectively transform workflows and decision-making.
Consider a financial institution that implements mid-tier projects for fraud detection, customer segmentation, and loan approval optimization. Individually, these projects solve distinct challenges, but together they compound ROI across the organization, building momentum for broader AI adoption. This approach establishes a sustainable foundation for long-term innovation.
3. Balancing Risk and Reward
Moonshots can carry high costs and uncertain outcomes, while low-level automations can offer limited differentiation. Automating repetitive tasks with rule-based systems or basic AI may improve efficiency, but these gains are incremental and rarely contribute to a lasting competitive edge. These solutions address narrow problems, lack adaptability, and often fail to align with broader strategic goals.
Mid-tier AI strikes the perfect balance: low-risk, scalable, and strategically impactful. These projects embed intelligence into workflows to solve complex, cross-functional challenges, delivering measurable ROI while fostering innovation. With mid-tier AI, businesses can achieve sustainable growth and stand out in competitive markets — without overextending resources.
Barriers to Scaling Mid-Tier AI
Despite their potential, mid-tier AI use cases face significant hurdles:
- Misaligned Priorities: Over-investing in moonshots or basic automations can leave mid-tier projects underfunded.
- Collaboration Gaps: Siloed communication between IT, data teams, and business units delays progress.
- Talent Shortages: Operational AI requires collaboration across skill sets, which can be challenging to manage.
Another significant barrier is measuring ROI. According to the "AI, Today" report by Dataiku and Databricks, many organizations struggle to track and validate the value of AI projects due to a lack of clear metrics. This challenge often leads to hesitation in scaling initiatives, including operational mid-tier AI projects, despite their potential for significant impact.
Dataiku — the Universal AI Platform — addresses these gaps by providing tools for metrics tracking, workflow visualization, and cross-departmental collaboration. These capabilities help organizations quantify and validate the impact of AI initiatives, ensuring mid-tier projects deliver measurable value.
Unlocking the Forgotten Middle With Dataiku
Supporting Operational AI That Drives ROI
Dataiku empowers organizations to operationalize mid-tier AI use cases by integrating with existing infrastructures, whether data resides in Snowflake, Databricks, legacy systems, or cloud environments like AWS, Azure, and Google Cloud. By working seamlessly across these environments, Dataiku avoids costly overhauls and enables businesses to focus on solving real-world pain points with greater efficiency.
To learn more about how Dataiku complements cloud platforms, explore our blog series:
- I Have AWS, Why Do I Need Dataiku?
- I Have Databricks, Why Do I Need Dataiku?
- I Have GCP, Why Do I Need Dataiku?
- I Have Microsoft, Why Do I Need Dataiku?
- I Have NVIDIA, Why Do I Need Dataiku?
- I Have Snowflake, Why Do I Need Dataiku?
For example, Air Canada transformed its post-campaign analysis process using Dataiku. By automating workflows and integrating with Snowflake and PowerBI, the airline achieved a 96% improvement over its prior marketing campaign analysis solution. This operational AI initiative not only saved significant time and costs but also freed up resources for teams to focus on more ambitious marketing innovations. As a foundational step, it sets the stage for building even more sophisticated AI-driven campaign strategies that could be considered moonshots in their own right.
Democratizing AI for Enterprise-Wide Impact
Mid-tier AI use cases thrive on collaboration between technical teams and business users. Dataiku’s no-code/low-code tools empower non-technical users to actively contribute, accelerating deployment and fostering innovation.
Using Dataiku, Air Canada reduced the time required to build predictive models from weeks to just hours. This democratization of AI allowed marketing teams to act quickly on data insights, improving both efficiency and business outcomes.
What’s great about [Dataiku] is it’s easy to look at the results in a quick way, make changes, and look at the results again. With business users that don’t understand really actually how it’s being done, but still understand the value, this was, really, really powerful and helped us to make a customer scoring model within two to three months that we’re going to now use on a day to day basis.— Hervé Riboulet, Director Cargo Analytics and CRM at Air Canada
3. Ready for Governance and Scalability
Governance and compliance are essential as organizations scale AI. Dataiku provides tools for model monitoring, traceability, and security, ensuring mid-tier AI projects meet business and regulatory standards.
For example, John Lewis Partnership (JLP) leveraged Dataiku to establish a strong governance framework for their AI models, ensuring compliance and scalability. By using Dataiku’s model monitoring, traceability, and security features, JLP was able to streamline their AI operations, enabling rapid deployment and maintaining regulatory standards while scaling their AI initiatives across the organization.
Stop Chasing Extremes, Start Unlocking the Middle
Mid-tier AI use cases aren't just a middle ground — they’re a key element of the engine of enterprise transformation. Businesses that focus on operational AI build a foundation for sustainable innovation, driving measurable ROI without the risks of moonshots or the limitations of basic automations.
Dataiku empowers organizations to turn operational AI into a strategic advantage. By unifying teams, streamlining processes, and embedding governance, Dataiku transforms the “forgotten middle” into a competitive edge. Businesses can stop chasing extremes and start scaling where it matters most, unlocking sustainable, enterprise-wide ROI.